name: PR Feedback
on:
pull_request:
types: [opened, synchronize, reopened]
branches-ignore: [ci-cd-maintenance]
permissions:
contents: read
pull-requests: write
checks: write
# Cancel duplicate runs
concurrency:
group: pr-feedback-${{ github.ref }}
cancel-in-progress: true
jobs:
# ==========================================
# Comprehensive Issue Detection & Reporting
# ==========================================
detect-and-report-issues:
name: Detect & Report Issues
runs-on: ubuntu-latest
if: github.event_name == 'pull_request'
steps:
- uses: actions/checkout@v4
with:
fetch-depth: 0 # Need full history for diff analysis
- uses: actions/setup-python@v5
with:
python-version: '3.12'
cache: 'pip'
- name: Install analysis tools
run: |
pip install --upgrade pip
pip install -e ".[dev,test]"
pip install ruff mypy bandit safety vulture
- name: Run comprehensive issue detection
id: issues
run: |
echo "## 🔍 PR Analysis Results" > pr-feedback.md
echo "" >> pr-feedback.md
echo "**PR**: #${{ github.event.number }} | **Commit**: \`${{ github.sha }}\`" >> pr-feedback.md
echo "" >> pr-feedback.md
# Initialize counters
format_issues=0
lint_issues=0
type_issues=0
security_issues=0
total_issues=0
# ===================
# 1. Format Issues
# ===================
echo "### 🎨 Code Formatting" >> pr-feedback.md
if ! ruff format --check . > format_output.txt 2>&1; then
format_issues=$(ruff format --check . 2>&1 | grep -c "would reformat" || echo "0")
echo "❌ **$format_issues file(s) need formatting**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see formatting issues</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
echo '```' >> pr-feedback.md
ruff format --check . 2>&1 | head -20 >> pr-feedback.md
echo '```' >> pr-feedback.md
echo "" >> pr-feedback.md
echo "**Fix**: Run \`ruff format .\` to fix all formatting issues" >> pr-feedback.md
echo "</details>" >> pr-feedback.md
total_issues=$((total_issues + format_issues))
else
echo "✅ **All files properly formatted**" >> pr-feedback.md
fi
echo "" >> pr-feedback.md
# ===================
# 2. Linting Issues
# ===================
echo "### 🔧 Code Linting" >> pr-feedback.md
if ! ruff check . --output-format=json > lint_output.json 2>/dev/null; then
lint_issues=$(jq length lint_output.json)
echo "❌ **$lint_issues linting issue(s) found**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see linting issues</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
# Group issues by severity and file
echo "#### 🚨 Issues by File:" >> pr-feedback.md
jq -r '
group_by(.filename) |
.[] |
"**" + .[0].filename + ":**\n" +
(map("- Line " + (.location.row | tostring) + ": " + .message + " (" + .code + ")") | join("\n"))
' lint_output.json >> pr-feedback.md
echo "" >> pr-feedback.md
echo "**Fix**: Run \`ruff check . --fix\` to auto-fix issues, or manually address remaining ones" >> pr-feedback.md
echo "</details>" >> pr-feedback.md
total_issues=$((total_issues + lint_issues))
else
echo "✅ **No linting issues found**" >> pr-feedback.md
fi
echo "" >> pr-feedback.md
# ===================
# 3. Type Issues
# ===================
echo "### 📝 Type Checking" >> pr-feedback.md
# First check for Python syntax errors (exclude import errors which are separate issue)
syntax_errors=0
for file in markitdown_mcp/*.py; do
if ! python -c "import ast; ast.parse(open('$file').read())" 2>/dev/null; then
syntax_errors=$((syntax_errors + 1))
fi
done
if [ "$syntax_errors" -eq 0 ]; then
# Only run mypy if files have valid syntax
if ! mypy markitdown_mcp --json-report mypy_report.json > mypy_output.txt 2>&1; then
type_issues=$(jq '.reports | keys | length' mypy_report.json 2>/dev/null || echo "1")
echo "❌ **Type checking issues found**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see type issues</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
echo '```' >> pr-feedback.md
cat mypy_output.txt | head -20 >> pr-feedback.md
echo '```' >> pr-feedback.md
echo "" >> pr-feedback.md
echo "**Fix**: Add proper type annotations and resolve type errors" >> pr-feedback.md
echo "</details>" >> pr-feedback.md
total_issues=$((total_issues + type_issues))
else
echo "✅ **Type checking passed**" >> pr-feedback.md
fi
else
echo "❌ **$syntax_errors Python syntax error(s) detected - cannot run type checking**" >> pr-feedback.md
type_issues=$syntax_errors
total_issues=$((total_issues + syntax_errors))
fi
echo "" >> pr-feedback.md
# ===================
# 4. Security Issues
# ===================
echo "### 🔒 Security Analysis" >> pr-feedback.md
if ! bandit -r markitdown_mcp/ -f json -o bandit_report.json > /dev/null 2>&1; then
high_issues=$(jq '[.results[] | select(.issue_severity == "HIGH")] | length' bandit_report.json)
med_issues=$(jq '[.results[] | select(.issue_severity == "MEDIUM")] | length' bandit_report.json)
if [ "$high_issues" -gt 0 ]; then
echo "🚨 **$high_issues HIGH severity security issue(s) - MUST FIX**" >> pr-feedback.md
security_issues=$high_issues
elif [ "$med_issues" -gt 0 ]; then
echo "⚠️ **$med_issues MEDIUM severity security issue(s) - Review recommended**" >> pr-feedback.md
security_issues=$med_issues
fi
if [ "$security_issues" -gt 0 ]; then
echo "" >> pr-feedback.md
echo "<details><summary>Click to see security issues</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
jq -r '
.results[] |
select(.issue_severity == "HIGH" or .issue_severity == "MEDIUM") |
"**" + .filename + ":" + (.line_number | tostring) + "**\n" +
"- **Issue**: " + .issue_text + "\n" +
"- **Severity**: " + .issue_severity + "\n" +
"- **Rule**: " + .test_name + "\n"
' bandit_report.json >> pr-feedback.md
echo "</details>" >> pr-feedback.md
total_issues=$((total_issues + security_issues))
fi
else
echo "✅ **No security issues detected**" >> pr-feedback.md
fi
echo "" >> pr-feedback.md
# ===================
# 5. Test Coverage Analysis
# ===================
echo "### 📊 Test Coverage Analysis" >> pr-feedback.md
# Check if tests directory exists and has test files
if [ -d "tests/unit" ] && [ "$(find tests/unit -name '*.py' | wc -l)" -gt 0 ]; then
if pytest tests/unit/ --cov=markitdown_mcp --cov-report=json --cov-report=term > coverage_output.txt 2>&1; then
coverage=$(jq -r '.totals.percent_covered' coverage.json)
coverage_int=${coverage%.*}
if [ "$coverage_int" -lt 80 ]; then
echo "❌ **Coverage $coverage% is below 80% requirement**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see coverage details</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
echo '```' >> pr-feedback.md
python -c 'import json; data=json.load(open("coverage.json")); files=data["files"]; uncovered=[(f,files[f]["summary"]["percent_covered"]) for f in files if files[f]["summary"]["percent_covered"]<80]; [print("Files below 80% coverage:"), [print(f"{f}: {p:.1f}%") for f,p in sorted(uncovered, key=lambda x: x[1])]] if uncovered else None' >> pr-feedback.md
echo '```' >> pr-feedback.md
echo "</details>" >> pr-feedback.md
total_issues=$((total_issues + 1))
else
echo "✅ **Coverage $coverage% meets 80% requirement**" >> pr-feedback.md
fi
else
echo "⚠️ **Tests failed to run - checking error details**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see test failures</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
echo '```' >> pr-feedback.md
cat coverage_output.txt | head -20 >> pr-feedback.md
echo '```' >> pr-feedback.md
echo "</details>" >> pr-feedback.md
fi
else
echo "⚠️ **No unit tests found in tests/unit/ directory**" >> pr-feedback.md
fi
echo "" >> pr-feedback.md
# ===================
# 6. Dead Code Detection
# ===================
echo "### 🧹 Dead Code Analysis" >> pr-feedback.md
if vulture markitdown_mcp/ --json > vulture_output.json 2>/dev/null; then
dead_code_count=$(jq length vulture_output.json)
if [ "$dead_code_count" -gt 0 ]; then
echo "⚠️ **$dead_code_count potential dead code item(s) found**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "<details><summary>Click to see potential dead code</summary>" >> pr-feedback.md
echo "" >> pr-feedback.md
jq -r '.[] | "**" + .filename + ":" + (.first_lineno | tostring) + "** - " + .message' vulture_output.json | head -10 >> pr-feedback.md
echo "</details>" >> pr-feedback.md
else
echo "✅ **No dead code detected**" >> pr-feedback.md
fi
else
echo "✅ **Dead code analysis completed**" >> pr-feedback.md
fi
echo "" >> pr-feedback.md
# ===================
# Summary & Recommendations
# ===================
echo "## 📋 Summary" >> pr-feedback.md
echo "" >> pr-feedback.md
if [ "$total_issues" -eq 0 ]; then
echo "🎉 **Excellent! No issues detected in this PR.**" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "All checks passed:" >> pr-feedback.md
echo "- ✅ Code formatting" >> pr-feedback.md
echo "- ✅ Linting rules" >> pr-feedback.md
echo "- ✅ Type checking" >> pr-feedback.md
echo "- ✅ Security analysis" >> pr-feedback.md
echo "- ✅ Test coverage" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "**Ready for review! 🚀**" >> pr-feedback.md
else
echo "⚠️ **Found $total_issues issue(s) that should be addressed:**" >> pr-feedback.md
echo "" >> pr-feedback.md
[ "$format_issues" -gt 0 ] && echo "- 🎨 Format: $format_issues file(s) need formatting" >> pr-feedback.md
[ "$lint_issues" -gt 0 ] && echo "- 🔧 Lint: $lint_issues issue(s)" >> pr-feedback.md
[ "$type_issues" -gt 0 ] && echo "- 📝 Types: Issues found" >> pr-feedback.md
[ "$security_issues" -gt 0 ] && echo "- 🔒 Security: $security_issues issue(s)" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "### 🔧 Quick Fix Commands:" >> pr-feedback.md
echo '```bash' >> pr-feedback.md
echo "# Fix formatting and auto-fixable linting issues" >> pr-feedback.md
echo "ruff format ." >> pr-feedback.md
echo "ruff check . --fix" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "# Run tests with coverage" >> pr-feedback.md
echo "pytest tests/unit/ --cov=markitdown_mcp --cov-report=term-missing" >> pr-feedback.md
echo "" >> pr-feedback.md
echo "# Check security" >> pr-feedback.md
echo "bandit -r markitdown_mcp/" >> pr-feedback.md
echo '```' >> pr-feedback.md
fi
echo "" >> pr-feedback.md
echo "---" >> pr-feedback.md
echo "*This analysis was automatically generated by the PR feedback workflow.*" >> pr-feedback.md
echo "*Report generated at $(date -u '+%Y-%m-%d %H:%M:%S UTC')*" >> pr-feedback.md
# Set outputs for other jobs
echo "total_issues=$total_issues" >> $GITHUB_OUTPUT
echo "format_issues=$format_issues" >> $GITHUB_OUTPUT
echo "lint_issues=$lint_issues" >> $GITHUB_OUTPUT
echo "security_issues=$security_issues" >> $GITHUB_OUTPUT
- name: Post comprehensive PR feedback
uses: marocchino/sticky-pull-request-comment@v2
with:
header: pr-analysis
recreate: true
path: pr-feedback.md
- name: Add PR labels based on issues
if: steps.issues.outputs.total_issues > 0
run: |
labels=""
# Add labels based on issue types
[ "${{ steps.issues.outputs.format_issues }}" -gt 0 ] && labels="$labels needs-formatting"
[ "${{ steps.issues.outputs.lint_issues }}" -gt 0 ] && labels="$labels needs-linting"
[ "${{ steps.issues.outputs.security_issues }}" -gt 0 ] && labels="$labels security-review"
# Add general labels
if [ "${{ steps.issues.outputs.total_issues }}" -gt 10 ]; then
labels="$labels needs-major-fixes"
elif [ "${{ steps.issues.outputs.total_issues }}" -gt 0 ]; then
labels="$labels needs-fixes"
fi
# Apply labels
for label in $labels; do
gh pr edit ${{ github.event.number }} --add-label "$label" || true
done
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
- name: Request changes if critical issues found
if: steps.issues.outputs.security_issues > 0
run: |
gh pr review ${{ github.event.number }} \
--request-changes \
--body "🚨 **Critical security issues detected!** Please review and fix the security issues identified in the automated analysis before this PR can be approved."
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
- name: Upload detailed reports
if: always()
uses: actions/upload-artifact@v4
with:
name: pr-analysis-reports
path: |
pr-feedback.md
lint_output.json
bandit_report.json
mypy_report.json
coverage.json
vulture_output.json
